80
Microfluidic bioassay to characterize parasitic nematode
phenotype and anthelmintic resistance
BAOZHEN CHEN 1 , ALEX DEUTMEYER 1 , JOHN CARR 1 , ALAN P. ROBERTSON 2 ,
RICHARD J. MARTIN 2 * and SANTOSH PANDEY 1
1 Department
2 Department
of Electrical and Computer Engineering, Iowa State University, Ames, IA 50011, USA
of Biomedical Sciences, Iowa State University, Ames, IA 50011, USA
(Received 3 May 2010; revised 4 June 2010; accepted 4 June 2010; first published online 21 July 2010)
SUMMA R Y
With increasing resistance to anti-parasitic drugs, it has become more important to detect and recognize phenotypes of
resistant isolates. Molecular methods of detecting resistant isolates are limited at present. Here, we introduce a microfluidic
bioassay to measure phenotype using parameters of nematode locomotion. We illustrate the technique on larvae of an animal
parasite Oesophagostomum dentatum. Parameters of sinusoidal motion such as propagation velocity, wavelength, wave
amplitude, and oscillation frequency depended on the levamisole-sensitivity of the isolate of parasitic nematode. The
levamisole-sensitive isolate (SENS) had a mean wave amplitude of 135 μm, which was larger than 123 μm of the levamisoleresistant isolate (LEVR). SENS had a mean wavelength of 373 μm, which was less than 393 μm of LEVR. The mean
propagation velocity of SENS, 149 μm s− 1, was similar to LEVR, 143 μm s− 1. The propagation velocity of the isolates was
inhibited by levamisole in a concentration-dependent manner above 0·5 μM. The EC50 for SENS was 3 μM and the EC50 for
LEVR was 10 μM. This microfluidic technology advances present-day nematode migration assays and provides a better
quantification and increased drug sensitivity. It is anticipated that the bioassay will facilitate study of resistance to other
anthelmintic drugs that affect locomotion.
Key words: microfluidics, bioassay, Oesophagostomum dentatum, levamisole resistance, phenotype.
I N T R OD UC T I O N
In animals, nematode parasites cause massive economic loss and welfare issues but in humans they
depress growth and learning at school in children,
they reduce work productivity in adults and help to
maintain poverty. Control of these infections relies
on regular chemotherapy but drug resistance is a
major concern in animals (Kaplan, 2004) and is
emerging in humans (Albonico et al. 2005). Early
detection of drug resistance is necessary for limiting
its spread. Although some molecular methods are
available for detecting a changed genotype associated with resistance (Schwenkenbecher et al. 2007;
Diawara et al. 2009), detection of resistance in
nematode parasites is mainly based on detecting a
changed phenotype. Many classes of anthelmintics
(haloxon, levamisole, monepantel, derquantel, ivermectin and piperazine) act on the neuromuscular
system of nematodes to inhibit locomotion (Martin,
1997; Martin et al. 2003). Drug resistance may be
associated with changes in the phenotype of this
locomotion because signalling-muscle-contraction
pathways are affected. In this paper, we describe a
method for characterizing the movement phenotype
* Corresponding author: Department of Biomedical
Sciences, Iowa State University, Ames, IA 50011, USA.
Tel: + 515 294 2470. Fax: + 515 294 3637. E-mail:
rjmartin@iastate.edu
of larvae of parasitic nematodes by quantitative
measurement of their locomotion in microfluidic
chambers. We show that phenotype varies with
parasite isolates and that the movement phenotype
is detectably different in a drug-resistant isolate.
Most nematodes rely on side-to-side sinusoidal
movement to ‘swim’ over solid surfaces. Their
mechanism of locomotion has intrigued researchers
for several decades and inspired a number of quantitative models to explain the nature of undulatory
motion (Gray, 1953; Gray and Lissmann, 1964;
McNeal, 2002; Lockery et al. 2008). In the model
nematode, C. elegans, locomotion experiments have
often been performed on agar plates where C. elegans
uses the surface tension of a thin surface water film
and its internal hydrostatic pressure to generate
the forward motion force (Pierce-Shimomura et al.
1999; Ryu and Samuel, 2002; Gray et al. 2004). The
oscillatory movement of the head is propagated
backwards along the rest of the body as sinusoidal
waves (McNeal, 2002; Gray et al. 2005). Recently,
microfluidics technology has enabled the fabrication
of a new class of devices such as micro-chambers,
mazes, wave sampler, and cylindrical posts to study
worm behaviour (Whitesides, 2006; Hulme et al.
2007; Qin and Wheeler, 2007; Rohde et al. 2007;
Lockery et al. 2008). Amongst other benefits, microfluidics technology offers experimental resolution
at the microscale level, thereby allowing the
Parasitology (2011), 138, 80–88. © Cambridge University Press 2010
doi:10.1017/S0031182010001010
Downloaded from https://www.cambridge.org/core, on subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0031182010001010
Microfluidics, phenotype and nematode locomotion parameters
81
levamisole at therapeutic doses (Martin et al. 2003).
The parameters under study were the velocity of
forward migration, amplitude and wavelength of the
sinusoidal motion, and the frequency of side-to-side
swimming action of the nematodes. We observed
clear differences between SENS and LEVR O.
dentatum larvae allowing the drug-resistant phenotype to be recognized in the absence of the anthelmintic. We characterized the anthelmintic dose
response for SENS and LEVR isolates and noticed
differences in the degree of their drug resistance.
MA TE R IAL S A ND ME THO D S
Fabrication of microfluidic devices
Fig. 1. Microfluidic chamber and larvae. (A) The fourchannel microfluidic device fabricated to study nematode
locomotion. There are vertical markers to help calibrate
the forward velocity. (B) Illustrations of a section of the
microfluidic device during a representative experiment.
Here, 2 isolates of Oesophagostomum dentatum (SENS and
LEVR) are put in alternating channels and videos were
recorded for data analysis. On the right, we show a
picture of the organism used in our experiments: O.
dentatum.
experimentalist to observe, track, and characterize
locomotion of a single worm over time (Whitesides,
2006). On-chip immobilization of nematodes on
microfluidic devices with high-resolution imaging
of its neurons and other cellular activity has also
generated great interest in the neuroscience community (Kerr et al. 2000; Nagel et al. 2005; Heng
et al. 2006; Hulme et al. 2007).
In this paper, we extend the application of
microfluidics technology to the study of parasitic
nematodes and to demonstrate differences in nematode phenotype. More importantly, the approach can
distinguish different isolates, levamisole–sensitive
and –resistant, of the same species. We designed
and fabricated microfluidic bioassays to record and
characterize the undulatory movement of larvae of
the pig parasite Oesophagostomum dentatum (Fig. 1A
and B). Two isolates of O. dentatum were tested:
levamisole-sensitive (SENS), which are sensitive
to the cholinergic anthelmintic levamisole and
levamisole-resistant (LEVR) which are resistant to
The microfluidic bioassays were designed to have 4
straight, parallel channels (width = 300 μm, length =
1 cm, height = 40 μm), each with its individual input
and output ports. The device dimensions can be
easily altered at the design stage to be adapted to the
selected species of nematodes under test, here O.
dentatum. The fabrication was made using standard
micromachining and soft lithography procedures
(Sia and Whitesides, 2003). The mask layout was
drawn in Adobe Photoshop and sent to an outside
vendor (Photoplotstore, Colorado Springs, CO, USA)
to generate the masks. A 40 μm-thick SU-8 photoresist (Microchem Corporation, Newton, MA, USA)
was cast on a bare silicon wafer by spin coating
(*2800 rpm). Standard photolithography was carried out to create the final SU-8 master mould. The
PDMS prepolymer (Sylgard 184 Silicone Elastomer
Kit, Dow Corning Corporation, Midland, MI) was
cast over the SU-8 mould and cured on a hot plate at
70 °C for 2 h. Subsequently, the PDMS microfluidic
devices were peeled off the mould and input/output
ports (2 mm diameter) were created using a sharp
needle. Air plasma was exposed to the PDMS
devices, which are then bonded to individual glass
cover-slips. The fabricated microfluidic bioassay is
shown in Fig. 1A and B, along with a recorded
picture of the nematodes moving through the sealed
channels.
Sample preparation
A 1% agar (Fisher Scientific) solution in water was
prepared. The mixture was put on a hot plate at
100 °C and stirred for 1 h. The molten agar was
allowed to cool down to 37 °C in a warm water bath
and 0·36 μl of agar was pipetted and filled in each of
the 4 channels of the microfluidic bioassay. Care was
taken to ensure that input ports of the channels are
not completely filled with agar or else, after a brief
time (2 min), the excess gel-like agar was scooped
and cleared out of the input ports. A small volume
(1 ml) of nematodes was pipetted and centrifuged to
increase the concentration of the worms (100/ml).
Downloaded from https://www.cambridge.org/core, on subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0031182010001010
Baozhen Chen and others
82
The nematodes were then added to the agar and
allowed to accommodate at 37 °C for 10 min.
Syringes were connected to the input ports of the
microfluidic bioassay using rubber tubing. Subsequently, the agar gel mixed with nematodes was
filled into the input ports of the 4 channels. After
5–15 min (the scouting time), the nematodes were
seen entering the channels from the input ports and
moving through the channel towards the output
ports.
Oesophagostomum dentatum larvae
Levamisole-sensitive (SENS) and levamisoleresistant (LEVR) O. dentatum were originally supplied by the Royal Veterinary and Agricultural
School, Frederiksberg, Copenhagen and then reproduced at 6–9 month intervals by passage in pigs
at Iowa State University, Ames, Iowa. The LEVR
was derived from SENS in the laboratory by using
increasing doses of levamisole for selection and
repeated passage. The L3 larvae isolates were maintained between passages in tap water refrigerated at
11 °C (changed every 2–4 months). LEVR-infected
hogs were treated orally with 8 mg/kg body weight to
maintain the selection pressure before collection of
eggs and preparation of larvae. The SENS L3 were
6 months old and the LEVR L3 were 5 months old
when used for our experiments. The larvae were
stored at 14 °C in tap water that was replaced at least
every 3 months.
Data capture and analysis
The nematodes were observed with a Leica MZ16
transmission stereomicroscope using magnifications
from 7·1 × to 230 ×. A QICAM 12-bit Mono
Fast 1394 Cooled digital camera interfaces with a
QCapture PRO software for capturing individual
digital images and digital image sequences. Digital
image sequences of nematode movement were recorded at 1 sec time-intervals with a resolution of
1392 × 1040 pixels and saved in uncompressed .avi
video format. These videos were later imported into
MATLAB where individual frames are analysed. An
actual distance to pixel distance ratio was created by
measuring the pixel width of a channel of known
width using the MATLAB Image Toolbox. This
ratio was used to translate pixel measurements to
actual distance for the calculation of nematode
velocity, amplitude, and wavelength. The MATLAB
Image Toolbox was use to measure pixel distance and
was able to accommodate the angle and direction of
nematode movement.
Calculation of locomotion parameters
Nematodes with unimpeded forward propagation
exhibit a consistent velocity over distances much
Fig. 2. Measurement of locomotion parameters.
(A) Illustrations of the SENS larvae for measurement of
the velocity (V), amplitude (A), and wavelength (λ) from
the recorded videos. Individual images were separately
analysed from a video and a MATLAB program was used
to accurately measure the parameters with a resolution of
0·1 mm. (B) A fourth parameter, oscillation frequency,
was calculated as the reciprocal of the total time-period
needed by a nematode to complete one sinusoidal wave
motion. The figure shows a series of time-lapsed images
of the larvae as it moved along the agar-filled channel and
the time-period periods separating T1, T2, T3 and T4 is
2 sec.
longer than their body length. Worm average velocity
was defined as the continuous forward linear distance
travelled by a worm’s head with respect to time. In
our experiments, average velocity over a length of the
channel was measured by obtaining the total number
of pixels travelled by the head of a nematode between
a certain number of image frames (minimum 20)
taken at 1 sec time-intervals. The total number of
pixels traversed between the beginning and the end
frames was calculated using QCapture PRO software
and translated to an actual distance using the previously mentioned ratio. This distance was then
divided by the total number of frames between the
beginning and end measurement positions to obtain
an average velocity in μm s−1.
Besides the average velocity, we measured parameters related to the sinusoidal body shape of nematodes during forward locomotion. The amplitude (A)
was measured from latitudinal peak to peak, while
wavelength (λ) was measured from longitudinal peak
to peak (Fig. 2A). The pixel distances of these
attributes were measured using the MATLAB Image
Toolbox. Pixel distance was converted to actual
distance in the units of micrometers using the previously mentioned ratio. The oscillation frequency
was calculated as the reciprocal of the total time
Downloaded from https://www.cambridge.org/core, on subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0031182010001010
Microfluidics, phenotype and nematode locomotion parameters
83
Fig. 4. Bar chart of the mean ±S.E. of wavelengths for
SENS and LEVR Oesophagostomum dentatum larvae
without any drug exposure, and SENS and LEVR
O. dentatum larvae with 1 mM levamisole. The SENS
wavelength was significantly (P < 0·02,*) less than LEVR
larvae in the absence of levamisole. Levamisole (1 μM) did
not have a significant effect on wavelength of either
isolate.
RE SU LTS
Amplitude measurements
Fig. 3. Amplitude measurements. (A) Top: 2 images of
SENS and LEVR Oesophagostomum dentatum. Below:
Amplitude histogram for SENS and LEVR O. dentatum
with no drug. (B) Bar charts of mean ±S.E amplitudes for
SENS and LEVR larvae with and without exposure to
1 mM levamisole. Significant differences, P < 0·0001 (***).
needed for a given worm to return to its original state
of sinusoidal pattern. A pictorial representation of the
oscillation frequency is shown in Fig. 2B.
These measurements of body parameters were only
performed on worms that were normal to the viewing
plane and clearly exhibit forward movement uninhibited by foreign objects. We ignored image
frame sequences where the worm under observation
touched other worms, the sidewalls of the channel or
any foreign particulates during its movement. This
prevented distortion of their natural amplitudes and
wavelengths and gave true information of their body
parameters. Furthermore, care was taken to use a
small concentration of worms (*100/ml) in each
channel so the chances of observing uninhibited
forward movement of nematodes in the microfluidic
bioassay increase.
Statistical analysis
Unpaired t-tests were used to estimate statistical
significance. Graph Pad V. 5.0 (GraphPad, San
Diego, CA, USA) was used to carry out these tests.
When we examined the amplitudes of the waves of
contraction, we found differences between the nematode isolates. Fig. 3A shows examples of the amplitude observed with sheathed SENS and LEVR
isolates of O. dentatum. Also shown are the distributions of SENS and LEVR amplitudes. SENS had
an amplitude of 135·2 ± 1·7 μm (mean, ±S.E., n = 32)
which was significantly (P < 0·001) larger than LEVR
(122·5 ± 1·8 μm, mean, ±S.E., n = 35).
When we measured amplitude of the SENS and
LEVR isolates in the presence of 1 μM levamisole, we
found that there was a significant reduction in the
amplitude of SENS (reduced to 121·3 ± 1·9 μm,
n = 38, P < 0·001) but not in LEVR (124·8 ± 1·9 μm,
n = 27, P = 0·39, Fig. 3B). A concentration of 1 μM
levamisole reduced the amplitude of SENS by 10·3%
but the LEVR amplitude did not decrease.
Wavelength measurements
SENS had a mean wavelength of 372·9 μm ± 7·1 μm
(mean, ±S.E., n = 32) and LEVR had a mean wavelength of 393·1 ± 4·6 μm (mean, ±S.E., n = 35) which
was significantly different (P = 0·019), Fig. 4. A concentration of 1 μM levamisole did not have a statistically significant effect on the wavelengths of either
of the isolates.
Tangential force measurements
An interesting aspect of the nature of nematode
locomotion involves the role of both amplitude and
wavelength, which had initially motivated us to
measure these body parameters. Gray (1953) found
Downloaded from https://www.cambridge.org/core, on subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0031182010001010
Baozhen Chen and others
84
that the maximum forward tangential force exerted
by a half-wave of the nematode body, FT, relates to its
body parameters and reported a detailed theoretical
analysis of undulatory motion. The normalized force
FT/Fo is expressed as (Gray, 1953; Lockery et al.
2008):
FT/Fo =
(2πA/λ)
λ[1 + (2πA/λ)2]1/2
(1),
in addition, is essentially proportional to A/λ. The
normalization factor Fo in equation (1) depends on
parameters such as surface tension, frictional force,
and coupling moment, which is assumed, unchanged
in our experiments (Gray, 1953). As shown in
Fig. 5A, the body of a nematode can be viewed as
different segments where each individual segment
applies a tangential force FT for the forward propulsive motion. The above equation also implies that
this tangential force increases with increase in amplitude and decrease in wavelength (Gray, 1953). The
data gathered for our amplitude and wavelength
analysis were used to obtain a set of A/λ quotients.
The mean ±S.E. A/λ quotient values of experiments on O. dentatum larvae are shown in Fig. 5B.
The mean of A/λ quotients for the SENS isolate was
0·366 ± 0·007 (mean ±S.E., n = 32), while that for the
LEVR isolate was 0·313 ± 0·005 (mean ±S.E., n = 35)
which is significantly smaller (P < 0·0001, unpaired
t-test). After exposure to 1 μM levamisole, the mean
A/λ quotient of SENS isolate showed a 13% decrease
to 0·317 ± 0·005 (mean ±S.E., n = 38) whereas the
LEVR isolate showed a 4·6% increase to 0·327 ±
0·006 (mean ±S.E., n = 27). The decrease in A/λ quotient for the SENS isolate was significant (P < 0·001)
while there was no statistically significant difference
in the A/λ quotient for the LEVR isolate with 1 μM
drug exposure (P > 0·05).
Following equation (1), the mean of the normalized tangential force for the SENS isolate was
2·49 ± 0·05 × 10−3 (mean ±S.E., n = 32) while that for
the LEVR isolate was significantly smaller (P < 0·001),
2·27 ± 0·03 × 10−3 (mean ±S.E., n = 35). After exposure to 1 μM levamisole, the mean normalized
force of the SENS isolate showed a 6·0% decrease
to 2·34 ± 0·04 × 10−3 (mean ±S.E., n = 38, P < 0·02)
whereas the LEVR isolate showed no significant
change and was 2·36 ± 0·04 × 10−3 (mean±S.E., n = 27,
P > 0·05). The percentage changes in the mean
normalized force among the O. dentatum isolates
follow a similar trend to the percentage changes in
their A/λ quotient values. This reiterates that the
tangential force needed for the forward movement
is directly proportional to the A/λ quotient and is
an important locomotion parameter for segregating
nematodes of different isolates and their response to
drugs.
Examination of the A/λ and normalized tangential
force results from the O. dentatum isolates leads
to interesting implications. The significantly lower
Fig. 5. Tangential forces and bar charts showing
mean ±S.E. values of A/λ of SENS and LEVR with and
without 1 μM levamisole. The tangential force, FT, is
proportional to A/λ. (A) The tangential force FT exerted
by each segment of the nematode’s body results in its
sinusoidal movement. (B) Bar charts of A/λ, the
amplitude to wavelength ratios, are shown for SENS and
LEVR without any levamisole and with 1 mM levamisole.
The SENS larvae had a significantly (P < 0·0001, ***)
greater amplitude to wavelength ratio than LEVR larvae.
A concentration of 1 μM levamisole significantly reduced
the SENS A/λ ratio but had no effect on LEVR.
values for the LEVR isolate, compared to the SENS
isolate, indicates that LEVR generates less forward
locomotive force, and suggests that LEVR muscles
produce less power. In the presence of levamisole,
both the A/λ quotient and the normalized force values
indicate that SENS muscles produced less force but
the effect on LEVR muscles was not significant. This
result is consistent with the larval migration studies
(Martin et al. 2003) where SENS is inhibited
more than LEVR by levamisole. We expect that the
tangential force exerted by the different segments of
the nematode body is proportional to the amount
of muscle contraction actively generating forward
movement. Since the cholinergic anthemintic levamisole affects body muscle contraction, it is not
surprising that we have observed distinct differences
in the locomotory parameters of SENS and LEVR
O. dentatum larvae when exposed to levamisole.
Forward velocity measurements
We compared the forward velocities of the O.
dentatum larvae with increasing concentrations of
levamisole, see Fig. 6. In the absence of levamisole,
Downloaded from https://www.cambridge.org/core, on subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0031182010001010
Microfluidics, phenotype and nematode locomotion parameters
Fig. 6. Velocity measurements. The measured forward
velocity for SENS (dotted) and LEVR (solid)
Oesophagostomum dentatum are plotted against the
increasing concentration of levamisole. At concentrations
higher than 1 μM, the velocity of SENS decreases more
rapidly than does that of LEVR. Line fitted by linear
interpolation between points. The EC50 for SENS isolate
was 3 μM while the EC50 for the LEVR isolate was 9 μM.
the SENS isolate had a velocity of 149·25 ± 6 μm s−1
(mean ±S.E., n = 31). LEVR had a velocity of
143 ± 4 μm s−1 (mean ±S.E., n = 37). The velocities
of both SENS and LEVR larvae were each measured
on the same 3 days. There was no statistically significant difference between the velocities of SENS
and LEVR isolates in the absence of levamisole.
Next, we tested the effect of different concentrations of levamisole on the SENS and LEVR
O. dentatum larvae. We measured the forward velocity of SENS and LEVR isolates when exposed to
levamisole. The velocities at various drug concentrations are plotted in Fig. 6. There was little decrease
in velocity for either SENS or LEVR isolates at a
levamisole concentration below 0·5 μM. However, the
SENS velocity drops clearly reduced with 1 μM while
a similar reduction in velocity for LEVR required
a levamisole concentration of 3 μM. The EC50 for
SENS isolate was 3 μM and the EC50 for LEVR
isolate was 10 μM (Fig. 6). Thus, we obtained a
resistance ratio of 3 between the two isolates as
measured from the forward velocities on the microfluidic device. From our previous migration studies
(Martin et al. 2003), the SENS EC50 was 26 μM and
the LEVR EC50 was 114 μM. This indicates that
the microfluidic device has approximately a 10-fold
better sensitivity compared to the migration assay
(Martin et al. 2003), but the resistance ratio between
SENS and LEVR is similar in both cases.
Oscillation frequency measurements
Oscillation frequency analysis was carried out to
further test measurements of velocity and wavelength
(Fig. 2B). As an additional and separate test,
velocity data and wavelength data were collected
independently from different videos. The predicted
85
frequency for all nematode sets was estimated by
dividing the mean velocity by mean wavelength.
The observed frequency was determined by observing
image frames, measuring the actual time taken by the
nematode larvae to return to its original state
(i.e. measure the time), and then taking the inverse
of this time. We found that the predicted frequency
values were close and not significantly different to the
observed frequency values. Interestingly, these observations were consistent with a classical equation of
physics (velocity =wavelength ×frequency) and this
equation was found to be valid when describing the
undulatory movement of nematodes.
The O. dentatum SENS larvae had an observed
mean frequency of 0·35 ± 0·02 s−1 (mean±S.E., n = 75)
which was not significantly differently from the
mean frequency of LEVR which was 0·38 ± 0·02 s−1
(mean ±S.E., n = 75). However, 1 μM levamisole reduced the frequency of SENS by 23% to 0·27 ±
0·02 s−1, n = 25 (P < 0·001) but produced little change
in the LEVR isolate (0·35 ± 0·02 s−1, P > 0·1).
DISCUSSION
Here we present a microfluidic bioassay with parallel
channels to observe and characterize nematode
locomotion parameters. We demonstrate that these
locomotion parameters are different for resistant and
sensitive isolates within the same species: SENS
and LEVR isolates of O. dentatum. Furthermore,
we showed the use of this microfluidic device for
observing drug effects on targeted nematode isolates
with better sensitivity than current migration assays.
Our approach is attractive because it allows us to
run multiples of experiments (each in an individual
channel) simultaneously and record the nematode
movement from all the channels. The low cost of
mass production of the device renders it disposable
(Hu, 2002) although it can be re-used multiple times.
The locomotion studies described in this paper
can be conducted in channels filled with water or
any suitable fluid. We preferred to use agar-filled
channels for the following reasons. (a) Agar provides
a better medium for the worms to push against as they
crawl and propel themselves forward. In a fluid
environment, the worms will swim and encounter
slips as they move forward. As such, a solid agar
medium would be closer to the natural soil habitat
of these worms. (b) In microfluidic channels, it is
difficult to maintain constant fluid pressure when
both sides of the channel are exposed to air. In our
initial experiments with water as a medium, the fluid
pressure in the channel interfered with the natural
movement of the nematodes. Such variation in fluid
pressure is not a concern when using agar or other
gel-like material as a medium in the channels.
Besides illustrating the difference in movement
phenotype between two drug-sensitive and drugresistant isolates, we think the usefulness of the
Downloaded from https://www.cambridge.org/core, on subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0031182010001010
Baozhen Chen and others
Fig. 7. Effect of higher concentrations of levamisole on
the locomotion parameters of SENS and LEVR
Oesophagostomum dendatum larvae. (A and B) A
concentration of 3 μM levamisole distorted the natural
sinusoidal movement of SENS larvae slowing velocity,
producing occasional curling and retraction from forward
movement. (C and D) At 3 μM levamisole had little effect
on the sinusoidal movement of LEVR larvae and their
locomotion parameters.
microfluidic device also lies in drug testing.
Currently, migration assays based on mesh structures
are used to characterize the effect of a drug on
nematodes (Martin et al. 2003). This testing is a slow
process, requiring many nematodes for a single test
and requiring higher concentrations of the drug.
With the microfluidic bioassays presented here, the
drug testing was done in parallel and each test was
completed within 45–60 min. This time, does not
include the extra time for data collection from the
video and analysis; we are now developing data
capture software and automatic analysis to speed this
process. However, compared to the binary migrated/
inhibited or hatch/not hatch measurements used in
larval migration or egg hatch assays, we are able to
measure the nematode velocity, wavelength, and
amplitude in the microfluidic bioassay.
The drug sensitivity is also significantly improved.
We have shown that the EC50 for the O. dentatum
larvae on the microfluidic bioassay is roughly 10
times smaller than that measured for O. dentatum
larvae on the migration assays. However, with higher
levamisole concentrations, the forward movement of
the larvae in the micro-channels becomes more
sluggish and discontinuous. Fig. 7 shows the upper
limit of the levamisole dose that was useful for
O. dentatum larvae in the microfluidic bioassay. At a
concentration of 2·2 μM levamisole, the SENS isolate
lost some of its ability to make sinusoidal movements
but LEVR retained this capacity. As shown in Fig. 6,
the forward velocity can still be measured in and
above this concentration of drug until the larvae is
completely immobile. This drug concentration limit
is reached at 10 μM of levamisole for the SENS isolate
86
and at 30 μM for the LEVR isolate. We used a typical
therapeutic dose of 8 mg/kg, given orally, to maintain
selection of LEVR during passage of O. dentatum
(see Materials and Methods section) through our
pigs. A dose of 8 mg/kg following oral administration
in hogs yields plasma concentrations of up to 7 μM
(De Baere et al. 2003). This concentration lies between the EC50 (3 μM) of SENS and the EC50 (10 μM)
of LEVR and is consistent with the therapeutic effect
of levamisole on SENS and its limited therapeutic
effect on LEVR. The effect of levamisole on larval
locomotory movements that we observed with our
microfluidic assay technique at lower concentrations
than with our larval migration inhibition assay
(Martin et al. 2003) may then relate more closely
to the therapeutically relevant effects on the adult
worm.
Other anthelmintic drug classes also affect locomotion: the macrocyclic lactones (ivermectin) produce paralysis of the body-wall muscle (Wolstenholme
and Rogers, 2005; Kotze et al. 2004); piperazine
inhibits body muscle as a GABA agonist (Martin,
1982); emodepside inhibits muscle, apparently increasing a Slo-1 potassium current (Guest et al.
2007). Even benzimidazoles, which bind β-tubulin,
affect larval motility (Kotze et al. 2004). We anticipate that all of these classes of anthelmintic drugs
could affect our phenotypic parameters of locomotion
in the resistant isolates, both in the absence and in the
presence of the anthelmintic.
Although our focus for this study was on the
detection and resolution of the phenotypes of parasitic larvae that are associated with anthelmintic
resistance, the methods are also likely to be of great
benefit for reverse genetic experiments where gene
knockdown, knockout or transgenic experiments
(Johnson et al. 2005) may produce subtle locomotory
phenotypes. We think the bioassay method is a
powerful tool that will also facilitate these studies.
The time required for data analysis can be reduced
and parameter extraction can be improved by using
automated nematode tracking software. Several
groups have developed software that automatically
calculates body parameters for C. elegans (Feng et al.
2004; Cronin et al. 2005; Chronis et al. 2007).
However, their software has to be customized to
measure the locomotion parameters as we have done.
Currently, we use MATLAB to calculate parameters
and characterize drug effects on nematode locomotion. It will speed analysis to automate the entire
data recording and analysis steps by adding a movable
test stage and developing a user-friendly program
for tracking nematodes in the microfluidic bioassay
under different drug concentrations.
We point out that like any bioassay; our method
has potential sources of error. A number of factors
need to be controlled for and can be minimized. For
example, movement of the larvae will be reduced by a
temperature reduction. This requires that collection
Downloaded from https://www.cambridge.org/core, on subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0031182010001010
Microfluidics, phenotype and nematode locomotion parameters
87
of the data for comparison be made at the same
temperature. We also found that as the age of the
larvae increased, there was a gradual decrease in the
velocity of movement; thus variation between isolates, even those that are drug sensitive may be
anticipated. However, we do expect that despite some
phenotypic variation between drug-sensitive isolates,
that they will show sensitivity to the drug under test,
in contrast to the resistance isolates. Despite potential
limitations, we think that these sources of error can
readily be controlled for and that our microfluidic
assay offers an improved and higher level of resolution.
Gray, J. (1953). Undulatory propulsion. Quarterly Journal
of Microscopical Science 94, 551–578.
Gray, J. and Lissmann, H. W. (1964). The locomotion of
nematodes. Journal of Experimental Biology 41, 135–154.
Gray, J. M., Hill, J. J. and Bargmann, C. I. (2005).
A circuit for navigation in Caenorhabditis elegans.
Proceedings of the National Academy of Sciences, USA
102, 3184–3191.
Gray, J. M., Karow, D. S., Lu, H., Chang, A. J.,
Chang, J. S., Ellis, R. E., Marletta, M. A. and
Bargmann, C. I. (2004). Oxygen sensation and social
feeding mediated by a C. elegans guanylate cyclase
homologue. Nature, London 430, 317–322.
Heng, X., Erickson, D., Baugh, L. R., Yaqoob, Z.,
Sternberg, P. W., Psaltis, D. and Yang, C. (2006).
Optofluidic microscopy—a method for implementing a
high resolution optical microscope on a chip. Lab Chip 6,
1274–1276.
Hu, S., Ren, X., Bachman, M., Sims, C. E., Li, G. P.
and Allbritton, N. (2002). Surface modification of poly
(dimethylsiloxane) microfluidic devices by ultraviolet
polymer grafting. Analytical Chemistry 74, 4117–4123.
Hulme, S. E., Shevkoplyas, S. S., Apfeld, J.,
Fontana, W. and Whitesides, G. M. (2007). A
microfabricated array of clamps for immobilizing and
imaging C. elegans. Lab Chip 7, 1515–1523.
Johnson, N. M., Behm, C. A. and Trowell, S. C. (2005).
Heritable and inducible gene knockdown in C. elegans
using Wormgate and the ORFeome. Gene 359, 26–34.
Kaplan, R. M. (2004). Drug resistance in nematodes of
veterinary importance: a status report. Trends in
Parasitology 20, 477–481.
Kerr, R., Lev-Ram, V., Baird, G., Vincent, P.,
Tsien, R. Y. and Schafer, W. R. (2000). Optical
imaging of calcium transients in neurons and pharyngeal
muscle of C. elegans. Neuron 26, 583–594.
Kotze, A. C., Clifford, S., O’Grady, J., Behnke, J. M.
and McCarthy, J. S. (2004). An in vitro larval
motility assay to determine anthelmintic sensitivity
for human hookworm and Strongyloides species.
American Journal of Tropical Medicine and Hygiene 71,
608–616.
Lockery, S. R., Lawton, K. J., Doll, J. C., Faumont, S.
and Coulthard, S. M. (2008). Artificial dirt:
Microfluidic substrates for nematode neurobiology and
behavior. Journal of Neurophysiology 99, 3136–3143.
Martin, R. J. (1982). Electrophysiological effects of
piperazine and diethylcarbamazine on Ascaris suum
somatic muscle. British Journal of Pharmacology 77,
255–265.
Martin, R. J. (1997). Modes of action of anthelmintic
drugs. Veterinary Journal 154, 11–34.
Martin, R. J., Bai, G., Clark, C. L. and Robertson, A. L.
(2003). Methyridine (2-[2-methoxyethyl]-pyridine])
and levamisole activate different ACh receptor
subtypes in nematode parasites: a new lead for
levamisole-resistance. British Journal of Pharmacology
140, 1068–1076.
McNeal, R. (2002). Locomotion. The Biology of
Nematodes (ed. Lee, D. L.), pp. 345–352. Taylor and
Francis, London, UK. 345.
Nagel, G., Brauner, M., Liewald, J. F., Adeishvili, N.,
Bamberg, E. and Gottschalk, A. (2005). Light
activation of channelrhodopsin-2 in excitable cells of
FI NA NC IA L S U P P O R T
The project was supported by Grant Number R 01 AI
047194 from the National Institute of Allergy and
Infectious Diseases to R.J.M., an Iowa Center for
Neurotoxicology grant to A.P.R. and Iowa State College
of Engineering funding to S.P. The content is solely the
responsibility of the authors and does not necessarily
represent the official views of the National Institute of
Allergy and Infectious Diseases of the National Institutes
of Health.
R EFE R EN C ES
Albonico, M., Wright, V., Ramsan, M., Haji, H. J.,
Taylor, M., Savioli, L. and Bickle, Q. (2005).
Development of the egg hatch assay for detection of
anthelmintic resistance in human hookworms.
International Journal for Parasitology 35, 803–811.
Chronis, N., Zimmer, M. and Bargmann, C. I. (2007).
Microfluidics for in vivo imaging of neuronal and
behavioral activity in Caenorhabditis elegans. Nature
Methods 4, 727–731.
Cronin, C. J., Mendel, J. E., Mukhtar, S., Kim, Y. M.,
Stirbl, R. C., Bruck, J. and Sternberg, P. W. (2005).
An automated system for measuring parameters of
nematode sinusoidal movement. BMC Genetics 6, 1–6.
De Baere, S., Cherlet, M., Croubels, S., Baert, K.
and De Backer, P. (2003). Liquid chromatographic
determination of levamisole in animal plasma: ultraviolet
versus tandem mass spectrometric detection. Analytical
Chimica Acta 483, 214–224.
Diawara, A., Drake, L. J., Suswillo, R. R., Kihara, J.,
Bundy, D. A. P., Scott, M. E., Halpenny, C.,
Stothart, J. R. and Pritchard, R. K. (2009). Assays to
detect beta-tubulin codon 200 polymorphism in
Trichuris trichiura and Ascaris lumbricoides. PLoS
Neglected Tropical Diseases 3, e397.
Feng, Z., Cronin, C. J., Wittig, J. H., Sternberg, P. W.
and Schafer, W. R. (2004). An imaging system for
standardized quantitative analysis of C. elegans behavior.
BMC Bioinformatics 5, 115.
Guest, M., Bull, K., Walker, R. J., Amliwala, K.,
O’Connor, V., Harder, A., Holden-dye, L. and
Hopper, N. A. (2007). The calcium-activated channel,
SLO-1, is required for the action of the novel
cyclo-octadepsipeptide anthelmintic, emodepside,
in Caenorhabditis elegans. International Journal for
Parasitology 37, 1577–1588.
Downloaded from https://www.cambridge.org/core, on subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0031182010001010
Baozhen Chen and others
Caenorhabditis elegans triggers rapid behavioral
responses. Current Biology 15, 2279–2284.
Pierce-Shimomura, J. T., Morse, T. M. and
Lockery, S. R. (1999). The fundamental role of
pirouettes in Caenorhabditis elegans chemotaxis.
Journal of Neuroscience 19, 9557–9569.
Qin, J. and Wheeler, A. R. (2007). Maze exploration and
learning in C. elegans. Lab Chip 7, 186–192.
Rohde, C. B., Zeng, F., Gonzalez-Rubio, R., Angel, M.
and Yanik, M. F. (2007). Microfluidic system for
on-chip high-throughput whole-animal sorting and
screening at subcellular resolution. Proceedings of the
National Academy of Sciences, USA 104, 13891–13895.
Ryu, W. S. and Samuel, A. D. (2002). Thermotaxis in
Caenorhabditis elegans analyzed by measuring responses
to defined thermal stimuli. Journal of Neuroscience 22,
5727–5733.
88
Sia, S. K. and Whitesides, G. M. (2003). Microfluidic
devices fabricated in poly(dimethylsiloxane) for
biological studies. Electrophoresis 24, 3563–3576.
Schwenkenbecher, J. M., Albonico, M., Bickle, Q.
and Kaplan, R. M. (2007). Characterization of
beta-tubulin genes in hookworms and investigation of
resistance-associated mutations using real-time PCR.
Molecular and Biochemical Parasitology 156, 167–174.
Whitesides, G. M. (2006). The origins and future of
microfluidics. Nature, London 442, 368–373.
Wolstenholme, A. J. and Rogers, A. T. (2005).
Glutamate-gated chloride channels and the mode of
action of the avermectin/milbemycin anthelmintics.
Parasitology 131, S85–S95.
Zhang, Y., Lu, H. and Bargmann, C. I. (2005).
Pathogenic bacteria induce aversive olfactory learning in
Caenorhabditis elegans. Nature, London 438, 179–184.
Downloaded from https://www.cambridge.org/core, on subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0031182010001010